View previous topic :: View next topic |
Author |
Message |
steveDoyle
Joined: 04 Sep 2009 Posts: 108 Location: Manchester
|
Posted: Mon Dec 21, 2020 3:34 pm Post subject: Active DO loop index altered |
|
|
Hi
Occasionally i get the "Active Do loop index altered" message when i'm debugging. Needless to say that the line of code is doing no such thing (i.e. an output statement) .
It only seem to appear when only some of the code is complied in debug mode i.e.
ftn95 test1.for /debug
ftn95 big_code.for
ftn95 reliable_code.for /debug
slink test1,big_code,reliable_code
The calling sequence is of the form Test1->big_code-> reliable_code
If this routine was in reliable_code.for
subroutine got_here(cbuff)
character*(*) cbuff
write(*,*,err=99) cbuff
99 continue
end
it would fail with "Active DO loop index altered" on the write statement
I haven't been able to reproduce the effect with a small example. Compiling without the debug option and the code works as expected
I understand that it maybe hard to track down without a reproducible example but an awareness of the issue maybe helpful to you
steve |
|
Back to top |
|
|
Robert
Joined: 29 Nov 2006 Posts: 446 Location: Manchester
|
Posted: Mon Dec 21, 2020 3:39 pm Post subject: |
|
|
Which version of the compiler are you using? |
|
Back to top |
|
|
steveDoyle
Joined: 04 Sep 2009 Posts: 108 Location: Manchester
|
Posted: Mon Dec 21, 2020 3:46 pm Post subject: |
|
|
Robert
ftn95 v8.70.0
Win xp 32
I have seen this message in previous versions and other platforms (win 8/10
64bit)
steve |
|
Back to top |
|
|
Robert
Joined: 29 Nov 2006 Posts: 446 Location: Manchester
|
Posted: Mon Dec 21, 2020 4:15 pm Post subject: |
|
|
I presume you are using checkmate. You mentioned /debug but that doesn't check for the do loop being modified. |
|
Back to top |
|
|
steveDoyle
Joined: 04 Sep 2009 Posts: 108 Location: Manchester
|
Posted: Mon Dec 21, 2020 4:20 pm Post subject: |
|
|
Robert
/full_debug
steve |
|
Back to top |
|
|
PaulLaidler Site Admin
Joined: 21 Feb 2005 Posts: 7931 Location: Salford, UK
|
Posted: Mon Dec 21, 2020 4:57 pm Post subject: |
|
|
steve
To make any progress, we would need some sample code that illustrates the failure.
If the failure concerns /FULL_DEBUG then you could switch it off before defining the routine...
options(-full_debug)
subroutine got_here(cbuff)
......
end
options(full_debug) |
|
Back to top |
|
|
steveDoyle
Joined: 04 Sep 2009 Posts: 108 Location: Manchester
|
Posted: Mon Dec 21, 2020 5:51 pm Post subject: |
|
|
Hi Paul
thanks for the quick response.
I understand these issues are next to impossible to resolve without a reproducible example. I will do some experimentation over the holiday period to see if i can isolate the problem
regards
steve |
|
Back to top |
|
|
StamK
Joined: 12 Oct 2016 Posts: 159
|
Posted: Mon Jan 04, 2021 1:28 am Post subject: |
|
|
Just wanted to add that I too have noticed this bug when running with undef. |
|
Back to top |
|
|
PaulLaidler Site Admin
Joined: 21 Feb 2005 Posts: 7931 Location: Salford, UK
|
Posted: Mon Jan 04, 2021 10:01 am Post subject: |
|
|
StamK
Can you demonstrate this bug with a short sample program? |
|
Back to top |
|
|
|